5104 B. Stothers, and A. L. Ternay, Jr.. J. Am. Chem. Soc.. 92,5912 (1970). (28) Isomer assignment was made on the basis of these values, together with the fact that Eu(fod)s causes substantially greater downfield shifts of Hd8in one isomer (assigned trans). (29) An additional possibility is a rapid boat-to-boat inversion which could produce a comparable homoallylic coupling constant. However, if one substituent can be accommodated by a planar 1,4-cyclohexadiene (Le., compound 2), presumably a second could be accommodated similarly, provided it is on the other side of the ring.
(30) M. E. Kuehne and 8. F. Lambert, d Am. Chem. SOC.,81,4278 (1959); Org. Synth., 43,23 (1963). (31) Attempts at complete NMR analysis with shift reagents were unsuccessful with this compound. (32) J. C. Marshall, K. C. Erickson. and T. K. Folsom, Tetrahedron Len., 401 1 (1970). (33) Structure was assigned to this dihydro isomer, since only one bridgehead proton showed significant coupling and there are only two vinylic protons which are shifted downfield. In addition, uv shows conjugation.
Proton Magnetic Shielding Tensors from Spin-Rotation Measurements on H2CO and NH3 S. G. Kukolich Contributionfrom the Department of Chemistry, The University of Arizona, Tucson, Arizona 85721. Received February 7, 1975
Abstract: The spin-rotation tensor for HzCO was obtained from very high resolution beam maser measurements of rotational transitions. Spin-rotation tensor elements are Ma, = -4.12 & 0.36 kHz, M b b = 2.45 & 0.32 kHz, and M,, = -2.10 f 0.34 KHz in the principal axis system. These results are combined with previous molecular orbital calculations of diamagnetic shielding to obtain total absolute shielding tensor elements of uaa = 15.9, Ubb = 40.7, and ucc= -1.7 pprn and an average Spin-rotation tenshielding of 18.3 ppm with uncertainties of &2 ppm or better. A similar analysis is carried out for ”3. sors and shielding tensors are given for proton and nitrogen sites with estimated accuracy better than I ppm. For NH, U,;(H) = 26.5 pprn and C(H) = 31.2 ppm. The absolute shielding scale is discussed.
The average of the diagonal elements of the magnetic shielding tensors, more commonly known as the “chemical shift”, has been extremely useful in structure determinations and characterizing particular types of bonding in functional groups. Since there are relatively large changes in shielding as the field direction is changed relative to molecular axes, a better understanding of magnetic shielding can be obtained from the shielding tensor. Shielding tensors have been obtained previously by orienting molecules in liquid crystals,’ using multiple-pulse techniques on single crystals,* or molecular beam Zeeman measurement^.^.^ Relatively few experiments have yielded accurate shielding tensors or quadrupole coupling tensors using liquid crystal techniques. Results which are in good agreement with other measurements were obtained for linear or symmetric top molecules. Difficulties appear to arise in establishing the order parameter. This problem is much more serious for asymmetric tops since the “orientation” axis is often not known and two-order parameters must be specified. Excellent results have been obtained from the multiple pulse measurements for larger molecules which may be frozen into rigid crystals and do not contain nuclei with quadrupole moments. Occasionally there are problems with this technique in determining the direction of shielding tensor axes relative to the bond directions. Direct measurements of shielding tensors for I9F and I3C have been obtained from high resolution molecular Zeeman measurements on molecular beams. These measurements are limited to a few small n;lecules with favorable magnetic and electric moments. i t was shown by RamseyS that the magnetic shielding tensor for diatomic molecules could be separated into diamagnetic and paramagnetic parts. The diamagnetic part is expressed as first-order matrix elements between groundstate wave functions. The paramagnetic part depends on second-order terms for the mixing of excited electronic states having nonzero orbital angular momentum with the ground state. This type of treatment for diatomics was extended to symmetric-top and asymmetric-top molecules by Flygare6 and coworkers. The diamagnetic terms may be Journal of the American Chemical Society
/
97:20
easily and accurately obtained from a b initio molecular orbital calculations. The paramagnetic terms are very difficult to calculate and accuracy is poor. The paramagnetic terms, however, are related to electron contributions to the spin-rotation tensor by a numerical constant. This suggests that accurate magnetic shielding tensors may be obtained by combining calculated diamagnetic terms with paramagnetic terms from spin-rotation measurements. This method was used previously for fluorine shielding tensors.’.* Formaldehyde rotational transitions were observed using a beam maser by Takuma, Shimizu, and S h i m ~ d a ,by ~ Thaddeus, Krisher, and Loubser,Io by Krupnov and SkvortS O VI, and ~ by Shigenari.’* More accurate measurements on J = 1 and J = 2 K-type doublets were reported by Tucker, Tomasevich, and Thaddeus. The resolution of the two-cavity maser ~pectrometerl~ using Ramsey’s method of separated oscillating fields is significantly better than the resolution of single cavity masers. Measurements of the 110-1 1 1 transition of H2CO using a two-cavity maser were reported previou~ly.’~
Experimental Section The two-cavity maser spectrometer was described previousI Y . ’ ~ For , ’ ~ the present measurements the beam was produced with a 0.1 mm diameter stainless steel nozzle. Formaldehyde vapor was obtained by heating a paraformaldehyde-silicone oil slurry. The upper state was focussed using a quadrupole lens. Two O F H C copper cavities excited in the TMolo mode were used to observe resonances. The cavity separation was 76 cm and the width of the central peak of the resonance pattern was 390 Hz. A “fast scan” recorder trace of all observed components is shown in Figure 1. The upper trace (a) is the “single-cavity” spectrum and the lower trace (b) was obtained using the two-cavity maser. Both spectra are shown as derivatives of the emission spectrum due to the frequency modulation with lock-in detection used in the apparatus.
- -
Spin-Rotation Tensor The strongest lines of the 211 212 transition and our previous data on the 110 1 I1 transition were analyzed to determine the spin-rotation tensor and spin-spin interac-
October 1, 1975
5705
A
Table I. Results of the Least-Squares Fit to J = 1 and J = 2 Hyperfine Structure in H,COQ ~~~
J
Fu
FL
Data
Calculated
Deviation
1
0 1 2 1 2 1 1 2 3 2
-18.67 -1.17 -0.46 3.96 6.46 11.08 -20.16 -7.88 0.95 10.82
-18.63 -1.44 -0.37 4.00 6.59 10.96 -20.07 -7.88 0.77 10.96
-0.04 0.27 -0.09 -0.04 -0.14 0.12 -0.09 0.00 0.19 -0.14
1 1 1 1 1 1 2 2 2 2
0 2 2 1 1 1 1 3 2
Q J = 1 line center at 4, 829, 659.9 kHz and J = 2 line center at 14,488,479.1 kHz. Frequencies in kHz relative to line centers. Standard deviation for the fit = 0.7 kHz. I
- 20
I -10
I 10
I 0
I
20
kHz
Table 11. Hyperfine Structure Parameters and Line Centers for H,CO
-
Parameter
Figure 1. Recorder traces of single-cavity (a) and two-cavity (b) spectra of the 211 212 transition in H2CO. The frequency scale is relative to the line center at 14,488,479. kHz.
Value, kHz -4.12 2.45 -2.10 17.65
Maa Mbb
Mcc
tion strength. The Hamiltonian for this problem was described p r e v i o ~ s l y . The ’ ~ results of a least-squares fit using the three diagonal elements of the spin-rotation tensor and spin-spin interaction strength as adjustable parameters are given in Table I. The standard deviation for the fit of 0.7 kHz was larger than the average statistical uncertainty in the individual like measurements of 0.13 kHz. Parameters determined from this analysis are given in Table 11. The spin-rotation parameters Mgg were in reasonably good agreement with previous work.13 The spin-spin interaction strength S = p1p2/r3 is smaller than the value 17.91 kHz calculated from the molecular geometry including the most important distortion term. Indicated error limits are one standard deviation only. The electronic contributions to the spin-rotation tensor are obtained by subtracting nuclear contributions from the total experimental spin-rotation tensor elements.
Magnetic Shielding Tensor The diagonal elements of the shielding tensor in the principal rotational axis system are expressed as the sum of the diamagnetic and paramagnetic contributions. u,,
=
uxxd
+
ux,p
The paramagnetic contribution to the xx component of the shielding tensor is directly related to the xx component of the spin-rotation tensor by subtracting a term representing nuclear contributions to the spin-rotation interaction M M,, - e2 Zn =1 C’ rn 7 b n 2 + zn2) 2mg1 G,, 2mc2 where M , is the proton mass, m is the electron mass, g1 is the nuclear g value, and M,JG,, is the ratio of spin-rotation strength to rotational constant in the x direction. The molecular geometry of H2CO is given by Oka.I6 The diamagnetic contributions to the shielding tensor were calculated by Neumann and Mosk~witz.~’ This was a HartreeFock, S C F calculation and appears to be the best ab initio calculation on H2CO. They also calculated nuclear contributions to spin-rotation tensor elements and we will use their results since they were obtained for the same geometry as electronic contributions to diamagnetic terms. Paramagnetic, diamagnetic, and total shielding tensor elements are listed in Table 111. For the paramagnetic term the listed uncertainty includes only the contribution from spin-rotation terms. There may be larger errors due to errors in geometry axxP
S
0.36 0.32 f 0.34 f 0.14 f f
1,,+ l,,center 4,829,659.9 ?: 0.2 2,, + 2,,center 14,488,479.1 t 0.3
Table 111. Paramagnetic (UxxP), Diamagnetic (uX,d), and Total (Uxx) Shielding Tensor Components along the Principal Rotational Axes in H,CW Axis a
b C
Av
axxp
-76.7 t 0.2 -54.0 t 1.4 -148.8 i 1.6 -69.3 1.2
*
OX,*
92.6 94.7 147.1 111.5
ax,
15.9 40.7 -1.7 18.3
a Units
are ppm. The listed uncertainty is only the contribution from uncertainty in the spin-rotation term.
but these should be cancelled by corresponding errors in diamagnetic terms. Diamagnetic shielding tensor components obtained from the semiempirical atom-dipole model18 are in reasonably good agreement with results from the a b initio c a l c ~ l a t i o n .We ’ ~ note from Table I11 that the largest shielding anisotropy is with respect to the b axis with 2Ubb - aaa - ucr = 67.2 ppm. This is well below the upper limit of 200 pprn established from high-resolution microwave Zeeman measurement^.'^ It is difficult to place error limits on the diamagnetic contribution axxdbut we would estimate that they are accurate to within a few parts per million. The diamagnetic shielding terms are only dependent on ground state wave functions and not very sensitive to the choice of basis set. It was pointed out by Rarnsey20 that there exists a linear relationship between ZM,,I,, and chemical shifts.20 This was extended by the Flygare-Goodisman2I model which relates the average spin-rotation contribution to shielding to the average shielding uaVsr(molecule) uav(freeatom) = u,,(molecule) where uavs‘ = ( M P / 6 m g 1 ) ZMxx/Gxx; for H2CO uaVsr = 0.8 ppm, and u,,(molecule) = 18.3. This would yield u,,(free atom) = 19.1 ppm. For hydrogen this value may be accurately calculated22 from ( l / r ) to be 17.75 ppm. An additional term included in a more recent treatmentI8 could possibly improve the model. The simplified form above is seen to be in error by only slightly more than 1 ppm.
+
NH3 Shielding Tensors A similar procedure may be used to calculate the proton and nitrogen shielding tensors in “3. High resolution Kukolich
/
Spin-Rotation Measurements on HzCO and N H J
5706 Table IV. Results of Least-Squares Fit of Observed Spin-Rotation Constants for Observed Transitions in lSNH,a J
K
CH(expt1)
C"(ca1cd)
CN(expt1) CN(calcd)
1 2 3 4 5 5
1 2 3 4 5 5
-18.362 -18.583 -18.711 -18.751 -18.770 -18.779
-18.382 -18.577 -18.674 -18.732 -18.771 -18.820
-9.496 -9.437 -9.404 -9.424 -9.418 -9.404
-9.485 -9.447 -9.428 -9.416 -9.408 -9.399
Table VII. Puamagnetic, Diamagnetic, and Total Shielding Tensor Components along Principal Rotational Axes in NH,,
Hydrogen
Axis
uggP(H)
Uggd(H)
ugg
z
-85.67 -55.10 -16.23 -93.98
112.18 88.60 57.39 119.82
26.51 33.50 41.16 25.84
Q
X
v,
Nitrogen
UggS*(H) 16.82 9.80 1.81 17.78
z
-118.27 355.97 237.70 -81.86 -75.96 353.92 277.96 -51.90 ,The Q axis represents the average of components perpendicular to the symmetry axis z so U, = (Uxx + Uy,,)/2, etc. Units ppm. Q
aExptl and calcd are experimental and calculated values in kHz. Table V. Results of Least-Squares Fit of Spin-Rotation Constants for Observed Transitions in I4NH,a
kHz. Rotational constants and structure were given by Benedict and P l ~ l e r . ~The ' best structure is probably the c' = 0 3 1 -17.829 -17.839 6.753 6.758 s t r ~ c t u r e , ~but ' it is very important to calculate nuclear 2 1 -17.939 -17.948 6.756 6.752 spin-rotation terms using the same structure as that used -17.992 4 2 -17.984 6.749 6.750 for calculating diamagnetic terms. High-accuracy Hartree-1 8.124 5 3 - 18.164 6.746 6.743 Fock calculations were reported by Laws, Stevens, and Lip-18.387 1 1 -18.379 6.729 6.729 scomb28 and later refined somewhat by Stevens.29W e will 2 2 -18.611 6.720 6.718 -18.606 use the results of the diamagnetic shielding calculations list- 18.7 16 3 3 -18.709 6.715 6.712 4 4 -18.781 -18.782 6.712 6.709 ed in Table VI of ref 28, equilibrium column, corresponding 5 5 -18.825 -18.826 6.698 6.706 to a structure with a proton (x = 1.754, z = 0.703 au) relative to the nitrogen atom. This structure is somewhat closer aValues in kHz to the e ~ p e r i m e n t aul ~=~ 0 structure than others used and a more complete set of proton shielding parameters is given. beam maser measurements on a large number of rotational The average absolute proton shielding is a ( H ) = 31.2 ppm states in I4NH3 and I5NH3 were reported by K ~ k o l i c h , ~ and ~ the average absolute nitrogen shielding 8(N) = 264.5 Kukolich and W o f ~ yand , ~ ~Ruben and K u k ~ l i c h Some .~~ ppm. The nitrogen shielding anisotropy is quite large with anomalies in the earlier work were removed in the theoretiuz, - uoa = -40.3 ppm. The shielding is greater in the cal analysis by Hougen26 which included a more complete plane of the molecule as observed previously for f l ~ o r i n e . ~ and careful theoretical treatment and inclusion of a new This is also the case for the protons in NH3 where the distortion effect on quadrupole coupling. The hydrogen and shielding anisotropy is uZz- uaa = 7 ppm. W e note that the nitrogen spin-rotation strengths for a given J , K rotational anisotropy between x and y directions is quite large but this state may be written on the form24 would not be observable in most experiments on ammonia due to the equivalence of axes perpendicular to the symmeCH(J,K) = - ( M Y x H M J L H ) / 2 - (2MZzH- M x x H try axis. In substituted amines these directions would no MyLH)K2/2J(J 1) ( M X x H- MLyH)a~f(-1)J+L/2 longer be equivalent. Here we have a(molecule) - 3' = 19.1 ppm, exactly the value obtained for H2CO and slightly Ch(J,K) = -MaaN - (Mzz\ - M,,')K2/J(J 1) larger than the free atom absolute shielding. The z axis is the C3 symmetry axis, the y axis is perpendicular to the N-H band and parallel to the hydrogen plane, Discussion and the x axis is parallel to the hydrogen plane and in the Absolute proton magnetic shielding values of 5 = 18.3 f plane of the z axis and N-H bond (see Figure 2, ref 24). 2.0 ppm and 5 = 31.2 f 1 .O ppm were obtained for H2CO Ma, = (M,, + MLL)/2and represents the component perand NH3 in the gas phase. These values, particularly the pendicular to the z axis. u represents the inversion quantum NH3 shielding, may be useful in establishing or checking an number. The results of fitting spin-rotation constants absolute shielding scale. The absolute shielding of H2 was C(J,K) for measured transitions to the above expressions determined to be 26.43 f 0.60 pprn by Myint et al.30 In the are given in Tables IV and V. The spin-rotation tensor elegas phase3I (i(NH3) = 4.25 relative to H2. The resulting ments are listed in Table VI. The term not listed in the tavalue a(NH3) = 30.7 f 0.6 is in agreement with the above bles is M X x H- MLyH= -28.98 f 0.03 based on four meavalue from the book by Carrington and M ~ L a c h l a nof~ ~ surements of K = l transitions. It is interesting to note that a(NH3) = 30.85 ppm. The same reference32 lists the I4N perpendicular and parallel components of hydrogen and niaverage shielding in NH3 as 266 ppm and we obtain trogen spin-rotation tensors are nearly equal but the per% . N ( N H ~=) 264.5 ppm. The accuracy of the results for the pendicular and parallel rotation constants and electronic average shielding provides some degree of confidence in the structure are quite different. accuracy of shielding tensors. Changes in proton chemical The paramagnetic shielding tensor elements for hydrogen shifts in ammonia due to isotopic substitution are very and nitrogen calculated from the above spin-rotation ten(>0.1 ppm) so we do not expect vibrational correcsors are listed in Table VII. The hydrogen spin-rotation tions to be important. tensor elements are M,, = 3.28, M y ) = 32.26, M,, = 19.01
+
+ +
+
Table VI. Spin-Rotation Tensor Elements Determined from Least-Squares Fits, Axisg a(perpendicu1ar) zloarallel) ~~
MaN 9.60 9.37
Mg*H
0.02 i 0.02 t
17.80 i 0.03 18.97 i 0.03
~
a Values
in kHz, Q perpendicular to symmetry axis and z parallel to symmetry axis.
Journal ofthe American Chemical Society
/ 97:20 / October I , 1975
Mg*N -6.764 -6.695
t f
MggH 0.005 i 0.005
17.73 i 0.02 19.05 i. 0.02
5707
Average magnetic shielding values for NH3 and H2CO have recently been calculated from molecular orbital theory by D i t ~ h f i e l d .These ~ ~ calculated values for hydrogen are a(H2CO) = 22.3 and C(NH3) = 33.6 ppm and for nitrogen u(NH3) = 267.8 ppm. The agreement with our above results is reasonably good in view of the difficulty of ab initio calculations of the paramagnetic term.
Acknowledgment. The writer wishes to thank Dr. Michael Barfield for helpful discussion on this work. The support of the National Science Foundation, Structural Chemistry Section, was greatly appreciated. References and Notes (1) P. K. Bhattacharyya and B. P. Dailey, J. Chem. Phys., 59, 5820 (1973). (2) S. Pausak, J. Tegenfeldt. and J. S. Waugh, J. Chem. Phys., 61, 1338 (1974): P. Van Hecke, T. C. Weaver, B. L. Neff, and J. S.Waugh, ibid., 60, 1668 (1974). and references cited therein. (3) F. H. De Leeuw and A. Dymanus. Chem. Phys. Lett., 7, 288 (1970). (4) S.G. Kukolich and A. C. Nelson, J. Chem. Phys., 56, 4446 (1972). (5) N. F. Ramsey, Phys. Rev., 76, 699 (1950); "Molecular Beams", Oxford University Press, London, 1956. Chapter 6. (6) W. H. Flygare. J. Chem. Phys., 41, 793 (1964): W. H ~ n e and r W. H. Flygare. ibid., 47, 4137 (1967). (7) S. G. Kukolich, T. H. S. Wang, and D. J. Ruben, J. Chem. Phys.. 58, 5474 (1973). (6) J. H. S.Wang and S.G. Kukolich, J. Am. Chem. Soc., 95, 4136 (1973). (9) H. Takuma. T. Shimizu, and K. Shimoda. J. Phys. SOC.Jpn. 14, 1595 (1959): 15, 2036 (1960); 16, 309 (1961).
(10) P. Thaddeus, L. C. Krisher, and J. H. N. Loubser. J. Chem. Phys., 40, 257 (1964). (11) A. F. Krupnov and V. A. Skvortsov, Sov. Phys. JETP (Engl. Trans/.), 40, 257 (1964). (12) T. Shigenari, J. Phys. Soc.Jpn., 23, 904 (1967). (13) K. D. Tucker, G. R. Tomasevich. and P. Thaddeus, Astrophys. J., 169, 429 (1971): 174, 463 (1972). (14) S.G. Kukolich, Phys. Rev., 138, A1322 (1965). (15) S.G. KukolichandD. J. Ruben, J. Mol. Spectrosc., 36, 130(1971). (16) T. Oka. J. Phys. SOC.Jpn., 15, 2279 (1960): 16, 1174 (1963). (17) D. B. Neurnann and J. W. Moskowitz, J. Chem. Phys., 50, 2216 (1969). (18) T. D. Gierke and W. H. Flygare, J. Am. Chem. Soc., 94, 7277 (1972). (19) S.G. Kukolich, J. Chem. Phys., 54, 8 (1971). (20) N. F. Ramsey, Am. Sci., 49, 509 (1961). (21) W. H. Flygare and J. Goodisman, J. Chem. Phys., 49,3122 (1968). (22) S. C. Wofsy. J. S. Muenter, and W. Klernperer, J. Chem. Phys., 55, 2014 (1971). (23) S.G. Kukolich, Phys. Rev., 138, A1322 (1965): 156, 63 (1967); 172, 59 (1968). (24) S.G. Kukolich and S.C. Wofsy, J. Chem. Phys., 52, 5477 (1970). (25) D. J. Ruben and S.G. Kukolich, J. Chern. Phys.. 61, 3780 (1974). (26) J. T. Hougen, J. Chem. Phys., 57, 4207 (1972). (27) W. S.Benedict and E. K. Plyler. Can. J. Phys., 35, 1235 (1957). (28) E. A. Laws, R. M. Stevens, and W. N. Libscornb, J. Chem. Phys., 56, 2029 (1972). (29) R. M. Stevens, J. Chem. Phys.. 61, 2086 (1974). (30) T. Myint. D. Kleppner, N. F. Rarnsey, and H. G. Robinson, Phys. Rev. Lett., 17, 405 (1966). (31) J. A. Poplg, W. G. Schneider, and H. J. Bernstein, "High-Resolution Nuclear Magnetic Resonance", McGraw-Hill, New York, N.Y., 1959, Chapter 5.2. (32) A. Carrington and A. D. McLachbn, "Introduction to Magnetic Resonance", Harper and Row, New York, N.Y., 1967, Chapter 5.4. (33) R. A. Bernstein and H. Batiz-Hernandez, J. Chem. Phys., 40, 3446 (1964). (34) R. Ditchfield, Mol. Phys., 27, 789 (1974).
High Resolution Nuclear Magnetic Resonance Studies of Chemical Reactions Using Flowing Liquids. Investigation of the Kinetic and Thermodynamic Intermediates Formed by the Attack of Methoxide Ion on 1-X-3,5-dinitrobenzenes' Colin A. Fyfe,*2 Michael Cocivera, and Sadru W. H. Damji Contribution from the Guelph- Waterloo Centrefor Graduate Work in Chemistry, Chemistry Department, University of Guelph, Guelph, Ontario, Canada N l G 2 W I , Received February 14, 1975
Abstract: High resolution nuclear magnetic resonance spectroscopy in a flowing system has been used to investigate both the kinetic and the thermodynamic distributions of the isomeric complexes formed by the attack of methoxide ion on 1-X-3.5dinitrobenzenes (where X = C N , CF3, COOCH3, COOCH2CH3) by the direct observation of these species at various reaction times in a flowing, chemically reacting system. In all cases a mixture of the two species is formed initially under kinetic control and the conversion from kinetically to thermodynamically controlled distributions occurs in less than 2 sec for all the compounds investigated. The results are compared with data on the reactions from previous uv-visible spectroscopic measurements and a reinterpretation of these data is given.
The various spectroscopic techniques used for kinetic measurements of chemical reactions all have their unique advantages and disadvantages. Thus, uv-visible spectroscopy has the advantage that it is a very sensitive technique and it is by far the most widely used method for making flow and stopped-flow kinetic measurements in reacting chemical systems. It does, however, have a severe disadvantage in that it is relatively nondiagnostic, the results leaving the identity of many observed species uncertain and in some cases depending on spectral assignments which may directly affect the interpretation of the kinetic data. Ir spectroscopy is less sensitive but is somewhat more diagnostic, depending on the presence of specific, often identifiable groupings of
atoms in the molecule. It is used only infrequently, however, mainly due to its lack of sensitivity. ESR combines a high degree of sensitivity with a considerable amount of diagnostic character but is obviously applicable only to a restricted class of compounds. However, it has found wide application for radical reactions and is often used where the species observed are in a flowing chemically reacting system, permitting the observation of relatively short-lived species. High-resolution nuclear magnetic resonance spectroscopy is much less sensitive than uv-visible spectroscopy (often by several orders of magnitude) but has a great advantage as a diagnostic tool and can provide much more information
Fyfe, Cocivera, Damji
/
Attack of Methoxide Ion on I -X-3.5-dinitrobenzenes
